Why deep understanding still matters in leadership in the age of AI

I only recently learned the term friction-maxxing. It means choosing to do things the harder way, like writing by hand instead of dictating, meeting in person instead of calling (which is tough for an introvert like me), or learning something from scratch instead of letting an algorithm do it. In many ways, it reflects something I’ve come to think of as leadership credibility.

I found the term in a Financial Times article, and honestly, my first reaction was relief, not curiosity. It turns out there’s a name for what I’ve been doing all along.

In the past few weeks, while working on an important project, I read nearly two hundred research papers. I made my own notes, took my time, and stayed with the material until I truly understood it.

Several people I respect told me I was wasting my time and should have automated the process or delegated the reading. They had a point about efficiency, but I’m not sure they were right about what’s valuable.

The question they raised is common in leadership: why do something yourself when a tool can do it faster? It’s a fair question, and I don’t ignore it. But I think it misses what I was really trying to do. I wasn’t just collecting information. I was building the kind of understanding you need to lead, guide, evaluate, challenge, and defend your decisions.

Conviction isn’t something you can automate or delegate.

There’s a phrase I keep coming back to: I can’t delegate what I don’t understand, let alone defend or explain it. This isn’t something I picked up from a self-help book – it’s just how I am.

I believed this long before the current debates about AI made it seem odd or admirable, depending on who you ask. I grew up before shortcuts were so common. We memorised things, made mistakes, fixed them, and learned by doing things badly at first, then better, and finally well if we kept at it. The effort wasn’t just part of learning; it was how we learned.

The FT article points out something important: researchers are warning about cognitive atrophy, which is the slow loss of our ability to think for ourselves when we let others –or technology – do the thinking before we try it ourselves.

This isn’t an argument against AI. I use it, my team uses it, and sometimes it’s the best tool for the job. But there’s a real difference between using a tool to speed up work you already understand and using it to skip the work of understanding altogether.

For me, the leadership part is what matters most. I can’t expect my team to do careful, thoughtful work if I haven’t shown I’m willing to do it myself. I can’t give real feedback, guide their thinking, or challenge them with any authority if I’m just reading summaries instead of the original material.

Leadership isn’t just about rank or making decisions. At its core, it’s about credibility, and you earn that in the hardest ways – by doing the long reading, making detailed notes, and being willing to deal with complexity instead of just turning it into a summary.

I also think about what I want to leave behind. I often wonder what it means to build something that lasts beyond a single job or project, and what it would mean for my colleagues to take something truly useful with them.

If they see a leader who hands off understanding as easily as tasks, I don’t think that’s a model worth following.

But if they see someone who does the hard work first and speaks from real knowledge – even if I’m wrong sometimes, but in an interesting way – maybe that’s a legacy worth having.

Ex nihilo nihil fit. Nothing comes from nothing.

I didn’t need the FT to tell me that. Still, it’s helpful to find out that a habit you’ve followed for thirty years, even if you couldn’t always explain it, turns out to be the right instinct after all.